Power resource scheduling and load balancing optimization method and system combined with big data

Abstract

The invention discloses a big data-combined power resource scheduling and load balancing optimization method and system, and the method comprises the steps: constructing a power demand prediction model, and obtaining a plurality of power demand prediction result sets in a future time period of each target region; inputting the real-time collected data into the power demand prediction model, and obtaining a prediction result with the highest matching degree; calculating the power load coefficient of each target region; constructing a regional load difference matrix, and obtaining an electric power resource calling-out region set and an electric power resource calling-in region set; dividing a scheduling period into a plurality of time periods, and performing power resource scheduling through incremental power transfer; and inputting the scheduled real-time acquired data into the power demand prediction model, and repeating the steps until the median value of the obtained regional load difference matrix is smaller than a scheduling threshold value. The method has the advantages that accurate power demand prediction is achieved through real-time data dynamic matching, and the power grid operation efficiency and stability are remarkably improved based on the load difference matrix and incremental scheduling optimization inter-region power distribution.

Description

Technical Field

[0001] The present invention relates to dynamic scheduling technology, and in particular to a method and system for optimizing power resource scheduling and load balancing in combination with big data. Background Art

[0002] Power resource scheduling and load balancing optimization are key technologies in modern power system operation, aiming to achieve a dynamic balance between electricity production, transmission, and consumption. With the large-scale integration of renewable energy sources such as wind and solar, the power system faces unprecedented challenges: these clean energy sources are characterized by significant intermittency and volatility, and their power generation output is highly uncertain due to weather conditions. Simultaneously, new trends are emerging on the power demand side, including the randomness of electric vehicle charging loads, the diversity of user electricity usage behavior, and sudden load fluctuations caused by extreme weather events. These factors, combined, have significantly increased the uncertainty of power grid operations.

[0003] Current methods for optimizing power resource scheduling and load balancing on the market primarily rely on simple load forecasting models and lack the deep integration and dynamic adjustment of real-time grid data. These methods typically employ static forecasting models that are unable to respond in real time to rapid changes in power demand and load, resulting in poor timeliness in scheduling decisions and difficulty optimizing the real-time configuration of power resources. Furthermore, when dealing with regional load differences in complex power systems, traditional methods often fail to effectively distinguish the actual demands of different regions, leading to wasted resources or uneven burdening of the grid. Furthermore, many methods lack continuous feedback and optimization mechanisms, making them unable to adapt to changes in grid conditions and demand, resulting in instability in long-term operation. Summary of the Invention

[0004] To improve existing power resource scheduling and load balancing optimization methods, this paper proposes a big data-based power resource scheduling and load balancing optimization method and system. This method uses big data analysis and an LSTM deep learning model to accurately predict power demand. It then dynamically adjusts power scheduling based on real-time grid data to optimize regional load distribution. It utilizes incremental power transfer and a greedy algorithm to improve scheduling efficiency, reduce resource waste, and ensure stable grid operation.

[0005] In order to achieve the above objects, the technical solution adopted by the present invention is:

[0006] The power resource scheduling and load balancing optimization method combined with big data includes:

[0007] Build an electricity demand forecast model based on historical electricity consumption data, and generate a set of multiple electricity demand forecast results for each target area in the future time period;

[0008] Based on the real-time data collected from the power grid nodes in each target area, it is input into the power demand forecasting model to obtain the forecast result with the highest matching degree among the various power demand forecast results.

[0009] The power load factor of each target area is calculated based on the power demand forecast result with the highest matching degree in each target area;

[0010] Based on the power load coefficient of each target area, a regional load difference matrix is constructed to obtain the set of power resource transfer-out areas and the set of power resource transfer-in areas;

[0011] Based on the obtained outbound and inbound area sets, the dispatch cycle is divided into multiple time periods, and power resources are dispatched through incremental power transfer;

[0012] Based on the real-time collected data of each target area after power resource scheduling, it is input into the power demand forecasting model and the above steps are repeated until the median of the obtained regional load difference matrix is less than the scheduling threshold.

[0013] Preferably, the step of constructing a power demand forecasting model based on historical power consumption data and generating a set of multiple power demand forecasting results for each target area in a future time period specifically includes:

[0014] Perform data preprocessing based on the acquired historical data and construct time series features and spatial features;

[0015] The acquired feature data is input into the LSTM-based deep model for training, and the accuracy of the model is evaluated through cross-validation method;

[0016] Based on the trained model, the electricity demand of each target area in different time periods is predicted, including short-term demand forecast and long-term demand forecast;

[0017] Based on the prediction results output by the model, a set of electricity demand prediction results with different confidence intervals is generated.

[0018] Preferably, the real-time data collection based on the power grid nodes in each target area, inputting the data into the power demand forecasting model, and obtaining the forecast result with the highest matching degree among the multiple power demand forecast result sets specifically includes:

[0019] Through the real-time monitoring system of the power grid nodes, real-time power data of each target area is collected, including current power consumption, grid frequency, voltage, and power transmission status;

[0020] Preprocessing is performed based on the collected real-time power data, and dynamic feature construction is performed;

[0021] The constructed feature vector is projected onto the pre-trained feature encoder, and dual matching is performed in both time and space dimensions, specifically by comparing the morphological similarity between the real-time sequence and the predicted curve through dynamic time warping, and by calculating the cosine similarity between the real-time topological feature vector and the predicted scene;

[0022] Based on the matching degree between real-time data and multiple prediction result sets, the prediction result with the highest similarity to the current power grid status is selected.

[0023] Preferably, the step of calculating the power load factor of each target area based on the power demand forecast result with the highest matching degree in each target area specifically includes:

[0024] Based on the power consumption data of each period in the time series data of the forecast result and the total number of the forecast period, the average load and maximum load in the period are obtained;

[0025] The calculated average load and maximum load are input into the load factor formula to obtain the power load factor;

[0026] Calculate the power load factor for each target area and obtain the power load factor time series dataset.

[0027] Preferably, constructing a regional load difference matrix based on the power load coefficient of each target area and obtaining a set of power resource transfer-out areas and a set of power resource transfer-in areas specifically includes:

[0028] Based on the obtained power load factor time series dataset, calculate the comprehensive load mean of all target areas;

[0029] Based on the comprehensive load mean, calculate the difference between the power load of each target area and the comprehensive load mean, and construct a regional load difference matrix;

[0030] Based on the difference values in the regional load difference matrix, a difference judgment threshold is set to divide and identify the outgoing and incoming areas;

[0031] Based on the degree of difference, the transferred-out areas and the transferred-in areas are ranked, with the transferred-in areas represented by positive numbers and the transferred-out areas by negative numbers, and a scheduling docking priority sequence is established.

[0032] Preferably, dividing the scheduling period into a plurality of time periods based on the obtained outgoing region set and incoming region set, and performing power resource scheduling through incremental power transfer specifically includes:

[0033] Divide the power dispatch cycle into hours, days, and weeks, and adjust it based on actual power demand;

[0034] Based on the established dispatch docking priority sequence, sort the incoming regions from largest to smallest according to the difference value, and sort the outgoing regions from smallest to largest, and select the two target regions with the closest absolute difference value;

[0035] Based on the divided dispatch period, the outgoing area will provide electricity during the period, and the incoming area will receive electricity during the period;

[0036] According to the load difference and power demand changes, the power dispatch amount in each period is dynamically adjusted through the greedy algorithm;

[0037] Real-time monitoring of power grid operation, actual load and dispatching effects in various regions is carried out, and the power demand forecasting model is retrained through the feedback mechanism to obtain stable forecasting results.

[0038] Furthermore, a power resource dispatching and load balancing optimization system combining big data is proposed, including:

[0039] Data acquisition module: The data acquisition module is mainly responsible for collecting key operating data such as power consumption, voltage, current, frequency, etc. of the power grid nodes in each target area in real time;

[0040] Data processing module: The data processing module is mainly used to clean, normalize and extract features of historical electricity consumption data to provide standardized input for subsequent modeling;

[0041] Power demand forecasting module: The power demand forecasting module is mainly used to make short-term and long-term forecasts of power demand in future time periods based on the LSTM model, and generate a set of forecast results under multiple confidence intervals;

[0042] Similarity matching module: The similarity matching module is mainly used to perform dual matching of time series and spatial features between real-time collected data and prediction sets, and select the optimal prediction result under the current state;

[0043] Load coefficient calculation module: The load coefficient calculation module is mainly used to calculate the average load and maximum load of each target area based on the selected prediction results, and obtain the load coefficient time series;

[0044] Regional difference analysis module: The regional difference analysis module is mainly used to build a regional load difference matrix, identify the areas where power resources are transferred out and in, and establish a dispatch priority list based on load differences;

[0045] Power resource scheduling module: The power resource scheduling module is mainly used to divide the scheduling cycle and use the greedy algorithm to formulate the optimal incremental power scheduling strategy to complete the transfer of power from the outgoing area to the incoming area;

[0046] Feedback module: The feedback module is mainly used to evaluate the scheduling effect and update the power forecast model based on the scheduling results and real-time operation data;

[0047] Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

[0048] Compared with the prior art, the advantages of the present invention are:

[0049] Through the use of an LSTM-based deep learning model and dynamic matching of real-time data, power demand is accurately predicted, ensuring the accuracy of scheduling decisions. Secondly, based on the power load factor and regional load difference matrix, the method can intelligently identify the regions where power is transferred in and out, and achieve precise power scheduling through incremental power transfer. By dividing the scheduling cycle into different periods and dynamically adjusting the amount of power transfer using a greedy algorithm, the flexibility and real-time performance of scheduling are further improved. Finally, the system's continuous feedback and retraining mechanism ensures the stability and long-term effectiveness of the power demand forecasting model. This method can significantly reduce the waste of power resources, optimize the distribution of power loads between regions, improve the operating efficiency and stability of the power grid, and promote green and sustainable energy management. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 A schematic diagram of the method proposed in the present invention;

[0051] Figure 2 This is a schematic diagram of the power demand forecasting model proposed in the present invention;

[0052] Figure 3 This is a schematic diagram of the prediction result matching proposed by the present invention;

[0053] Figure 4 This is a schematic diagram of obtaining the power load factor proposed by the present invention;

[0054] Figure 5 This is a schematic diagram of the division of incoming and outgoing transfer areas proposed by the present invention;

[0055] Figure 6 This is a schematic diagram of the power dispatch proposed by the present invention;

[0056] Figure 7 This is a diagram of the architecture of the electronic equipment in this solution;

[0057] Figure 8 This is a schematic diagram of the computer-readable storage medium structure in this solution. DETAILED DESCRIPTION

[0058] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0059] The power resource scheduling and load balancing optimization system combined with big data includes:

[0060] Data acquisition module: The data acquisition module is mainly responsible for collecting key operating data such as power consumption, voltage, current, frequency, etc. of the power grid nodes in each target area in real time;

[0061] Data processing module: The data processing module is mainly used to clean, normalize and extract features of historical electricity consumption data to provide standardized input for subsequent modeling;

[0062] Power demand forecasting module: The power demand forecasting module is mainly used to make short-term and long-term forecasts of power demand in future time periods based on the LSTM model, and generate a set of forecast results under multiple confidence intervals;

[0063] Similarity matching module: The similarity matching module is mainly used to perform dual matching of time series and spatial features between real-time collected data and prediction sets, and select the optimal prediction result under the current state;

[0064] Load coefficient calculation module: The load coefficient calculation module is mainly used to calculate the average load and maximum load of each target area based on the selected prediction results, and obtain the load coefficient time series;

[0065] Regional difference analysis module: The regional difference analysis module is mainly used to build a regional load difference matrix, identify the areas where power resources are transferred out and in, and establish a dispatch priority list based on load differences;

[0066] Power resource scheduling module: The power resource scheduling module is mainly used to divide the scheduling cycle and use the greedy algorithm to formulate the optimal incremental power scheduling strategy to complete the transfer of power from the outgoing area to the incoming area;

[0067] Feedback module: The feedback module is mainly used to evaluate the scheduling effect and update the power forecast model based on the scheduling results and real-time operation data;

[0068] Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

[0069] See Figure 1 As shown in FIG, the power resource scheduling and load balancing optimization method combined with big data includes:

[0070] Step 1: Build a power demand forecasting model based on historical power consumption data, and generate a set of multiple power demand forecast results for each target area in the future time period;

[0071] Step 2: Based on the real-time data collected from the power grid nodes in each target area, the data is input into the power demand forecasting model to obtain the forecast result that best matches the set of multiple power demand forecast results;

[0072] Step 3: Calculate the power load factor of each target area based on the power demand forecast result with the highest matching degree in each target area;

[0073] Step 4: Based on the power load coefficient of each target area, construct a regional load difference matrix to obtain the set of power resource transfer-out areas and the set of power resource transfer-in areas:

[0074] Step 5: Based on the obtained outbound and inbound area sets, the dispatch cycle is divided into multiple time periods, and power resources are dispatched through incremental power transfer;

[0075] Step 6: Based on the real-time collected data of each target area after power resource scheduling, input it into the power demand forecasting model and repeat the above steps until the median of the obtained regional load difference matrix is less than the scheduling threshold.

[0076] See Figure 2 As shown in the figure, a power demand forecasting model is constructed based on historical power consumption data, and a set of power demand forecast results for each target area in the future time period is generated, specifically including:

[0077] Perform data preprocessing based on the acquired historical data and construct time series features and spatial features;

[0078] The acquired feature data is input into the LSTM-based deep model for training, and the accuracy of the model is evaluated through cross-validation method;

[0079] Based on the trained model, the electricity demand of each target area in different time periods is predicted, including short-term demand forecast and long-term demand forecast;

[0080] Based on the prediction results output by the model, a set of electricity demand prediction results with different confidence intervals is generated.

[0081] Specifically, the LSTM architecture consists of three layers: the input layer accepts feature data of time steps, such as power demand at the previous moment, temporal features, spatial features, etc.; the LSTM layer uses multiple LSTM units to learn long-term dependencies; and the output layer predicts power demand at future moments.

[0082] The cross-validation method is used to evaluate the model performance to prevent overfitting, and MSE (mean square error) is used as the evaluation indicator. The formula is:

[0083]

[0084] Among them, yi is the actual value, is the predicted value, N is the number of historical data samples;

[0085] For short-term demand forecasting, the trained LSTM model is used to predict electricity demand in the next few hours or days. For long-term demand forecasting, external factors (such as seasonal changes, policy changes, etc.) and long-term trends need to be combined.

[0086] Based on the predicted value output by the model, multiple prediction models are obtained by resampling the training data multiple times. These prediction results are combined and calculated to obtain the confidence interval of the predicted value. The prediction results with different confidence intervals are generated according to the model output. The confidence interval formula is:

[0087]

[0088] in, is the predicted value, is the standard deviation of the prediction error, n is the number of samples, is the critical value of the standard normal distribution.

[0089] See Figure 3 As shown, based on the real-time data collected from the power grid nodes in each target area, it is input into the power demand forecasting model to obtain the forecast result with the highest matching degree among the multiple power demand forecast result sets, including:

[0090] Through the real-time monitoring system of the power grid nodes, real-time power data of each target area is collected, including current power consumption, grid frequency, voltage, and power transmission status;

[0091] Preprocessing is performed based on the collected real-time power data, and dynamic feature construction is performed;

[0092] The constructed feature vector is projected onto the pre-trained feature encoder, and dual matching is performed in both time and space dimensions, specifically by comparing the morphological similarity between the real-time sequence and the predicted curve through dynamic time warping, and by calculating the cosine similarity between the real-time topological feature vector and the predicted scene;

[0093] Based on the matching degree between real-time data and multiple prediction result sets, the prediction result with the highest similarity to the current power grid status is selected.

[0094] Specifically, the time series data is sliced through a sliding window to construct dynamic features between the current moment and the historical moment. The real-time data is projected into a pre-trained feature encoder that can map the input high-dimensional feature space to a low-dimensional space.

[0095] For each set of prediction results, calculate the similarity with the current real-time grid status, specifically including:

[0096] The morphological similarity between real-time sequence data and prediction result set data is obtained through dynamic time warping matching. The formula is:

[0097]

[0098] Among them, Q i For real-time sequence data, R j Aggregate data for prediction results;

[0099] When calculating topological cosine similarity, the real-time topological vector g t With the prediction scenario h (k) The matching calculation formula is:

[0100]

[0101] The formula considering the similarity of tidal flow direction is:

[0102]

[0103] Based on dual matching in time and space, the joint similarity score is calculated using the formula:

[0104]

[0105] Among them, the weight α+β+γ=1, which is obtained through optimization of historical events of the power grid.

[0106] See Figure 4 As shown, the power load factor of each target area is calculated based on the power demand forecast result with the highest matching degree in each target area, specifically including:

[0107] Based on the power consumption data of each period in the time series data of the forecast result and the total number of the forecast period, the average load and maximum load in the period are obtained;

[0108] The calculated average load and maximum load are input into the load factor formula to obtain the power load factor;

[0109] Calculate the power load factor for each target area and obtain the power load factor time series dataset.

[0110] Specifically, assuming that the power consumption in a certain period of time in the time series is P t (t represents the time step. In a certain period of time, the power consumption sequence is [P1, P2, ..., P n ], where n is the number of time steps in the period, and the average load can be calculated using the following formula:

[0111]

[0112] Among them, L avg is the average load during this period, P t is the electricity consumption at each time step;

[0113] The power load factor is used to measure power demand. It represents the ratio of maximum load to average load over a period of time. A smaller load factor means a greater power demand; a larger load factor means a smaller power demand. The formula is:

[0114]

[0115] Where LF is the load factor, L extmax is the maximum load during this period, L extavg is the average load during the period;

[0116] For multiple target regions, we can calculate the power load factor of each region according to the above method and generate a time series dataset of the load factor of each region, which is recorded as:

[0117] P extregion1 =[P 1,1 ,P 1,2 ,…,P 1,n ],P extregion2 =[P 2,1 ,P 2,2 ,…,P 2,n ],…

[0118] Among them, P extregion1 、P extregion2 Represent the electricity demand series of each region respectively. For each region, calculate the average load and maximum load of the region during the period, and then use the above formula to calculate the load factor. Finally, we can obtain the load factor time series dataset of each region.

[0119] See Figure 5 As shown in the figure, based on the power load coefficient of each target area, a regional load difference matrix is constructed to obtain the power resource outbound and inbound area sets, specifically including:

[0120] Based on the obtained power load factor time series dataset, calculate the comprehensive load mean of all target areas;

[0121] Based on the comprehensive load mean, calculate the difference between the power load of each target area and the comprehensive load mean, and construct a regional load difference matrix;

[0122] Based on the difference values in the regional load difference matrix, a difference judgment threshold is set to divide and identify the outgoing and incoming areas;

[0123] Based on the degree of difference, the transferred-out areas and the transferred-in areas are ranked, with the transferred-in areas represented by positive numbers and the transferred-out areas by negative numbers, and a scheduling docking priority sequence is established.

[0124] Specifically, the comprehensive load mean is obtained by taking a weighted average of the load factors of all regions. For the mth period, the calculation formula for the comprehensive load mean of all target regions is:

[0125]

[0126] in, is the average comprehensive load value in the mth period, is the load factor of the ith region in the mth period;

[0127] Calculate the difference between the power load of each target area and the comprehensive load average. The difference value represents the degree of deviation between the load factor of each area and the overall load average. For the i-th area, the difference value calculation formula for the m-th time period is:

[0128]

[0129] in, is the difference value of the i-th region in the m-th period, is the load factor of the ith region in the mth period, is the average comprehensive load value in the mth period;

[0130] Based on the difference values obtained for each target area, a regional load difference matrix is constructed. By setting a difference threshold, high-load and low-load areas are divided, and the incoming and outgoing areas are sorted according to the degree of regional load difference. The incoming and outgoing areas will be prioritized according to the size of their load differences. Positive numbers represent incoming areas, and negative numbers represent outgoing areas.

[0131] See Figure 6 As shown, based on the obtained outgoing area set and incoming area set, the scheduling period is divided into multiple time periods, and the power resource scheduling is performed through incremental power transfer, specifically including:

[0132] Divide the power dispatch cycle into hours, days, and weeks, and adjust it based on actual power demand;

[0133] Based on the established dispatch docking priority sequence, sort the incoming regions from largest to smallest according to the difference value, and sort the outgoing regions from smallest to largest, and select the two target regions with the closest absolute difference value;

[0134] Based on the divided dispatch period, the outgoing area will provide electricity during the period, and the incoming area will receive electricity during the period;

[0135] According to the load difference and power demand changes, the power dispatch amount in each period is dynamically adjusted through the greedy algorithm;

[0136] Real-time monitoring of power grid operation, actual load and dispatching effects in various regions is carried out, and the power demand forecasting model is retrained through the feedback mechanism to obtain stable forecasting results.

[0137] Specifically, the power dispatch cycle can be divided into hours, days, and weeks to adapt to power demand at different time scales. Within each cycle, the power dispatch amount will be optimized based on demand and the grid's carrying capacity. Hourly dispatch is suitable for short-term power load fluctuations and is usually used for peak-valley regulation. Daily dispatch is adjusted according to daily power demand and is suitable for daily power demand forecasting. Weekly dispatch is suitable for long-term load forecasting and takes into account seasonality, weekend effects, etc. The formula for the dynamic adjustment rule is:

[0138]

[0139] Among them, ΔT is the amplitude of the scheduling cycle adjustment, which is used to determine whether the scheduling cycle needs to be re-divided, and α is the sensitivity coefficient, which is used to control the sensitivity of the adjustment. is the predicted load value for period t, L t-1 is the actual load value in period t-1. When ΔT>5%, it triggers cycle re-division.

[0140] In each dispatch cycle, based on the load differences between regions, the outgoing regions provide electricity and the incoming regions receive electricity. If dispatch is required within a certain time period, the power distribution between the incoming and outgoing regions will be dynamically adjusted based on the load differences.

[0141] Using a greedy algorithm, the power dispatching amount is adjusted according to the difference in power demand in each period. By selecting the current optimal dispatching amount each time, the power demand in each cycle is reasonably met.

[0142] By monitoring the power demand and actual load in each region in real time and continuously adjusting the power dispatch strategy based on the feedback results, the power demand forecasting model needs to be retrained after each cycle to adapt to new demand changes.

[0143] Furthermore, the method according to the embodiment of the present application can also be used with the aid of Figure 7 The electronic device architecture shown in FIG. Figure 7As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, may store the power resource scheduling and load balancing optimization method and system combined with big data provided in this application. The electronic device 500 may also include a terminal interface 508. Of course, Figure 7 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 7 One or more components of an electronic device are shown.

[0144] Figure 8 This is a schematic diagram of the computer-readable storage medium structure provided by an embodiment of the present application. Figure 8 As shown, a computer-readable storage medium 600 according to one embodiment of the present application is shown. Computer-readable instructions are stored on the computer-readable storage medium 600. When the computer-readable instructions are executed by the processor, the power resource scheduling and load balancing optimization method and system combined with big data according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0145] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0146] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

[0147] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The power resource scheduling and load balancing optimization method combined with big data is characterized by: include: Build an electricity demand forecast model based on historical electricity consumption data, and generate a set of multiple electricity demand forecast results for each target area in the future time period; Based on the real-time data collected from the power grid nodes in each target area, it is input into the power demand forecasting model to obtain the forecast result with the highest matching degree among the various power demand forecast results. The power load factor of each target area is calculated based on the power demand forecast result with the highest matching degree in each target area; Based on the power load coefficient of each target area, a regional load difference matrix is constructed to obtain the set of power resource transfer-out areas and the set of power resource transfer-in areas; Based on the obtained outbound and inbound area sets, the dispatch cycle is divided into multiple time periods, and power resources are dispatched through incremental power transfer; Based on the real-time collected data of each target area after power resource scheduling, it is input into the power demand forecasting model and the above steps are repeated until the median of the obtained regional load difference matrix is less than the scheduling threshold.

2. The method for optimizing power resource scheduling and load balancing based on big data according to claim 1, characterized in that: The method of constructing a power demand forecasting model based on historical power consumption data and generating a set of multiple power demand forecasting results for each target area in the future time period specifically includes: Perform data preprocessing based on the acquired historical data and construct time series features and spatial features; The acquired feature data is input into the LSTM-based deep model for training, and the accuracy of the model is evaluated through cross-validation method; Based on the trained model, the electricity demand of each target area in different time periods is predicted, including short-term demand forecast and long-term demand forecast; Based on the prediction results output by the model, a set of electricity demand prediction results with different confidence intervals is generated.

3. The method for optimizing power resource scheduling and load balancing based on big data according to claim 1, characterized in that: The method of collecting data in real time based on the power grid nodes in each target area and inputting the data into the power demand forecasting model to obtain the forecast result with the highest matching degree among the multiple power demand forecast result sets specifically includes: Through the real-time monitoring system of the power grid nodes, real-time power data of each target area is collected, including current power consumption, grid frequency, voltage, and power transmission status; Preprocessing is performed based on the collected real-time power data, and dynamic feature construction is performed; The constructed feature vector is projected onto the pre-trained feature encoder, and dual matching is performed in both time and space dimensions, specifically by comparing the morphological similarity between the real-time sequence and the predicted curve through dynamic time warping, and by calculating the cosine similarity between the real-time topological feature vector and the predicted scene; Based on the matching degree between real-time data and multiple prediction result sets, the prediction result with the highest similarity to the current power grid status is selected.

4. The method for optimizing power resource scheduling and load balancing based on big data according to claim 1, wherein: The calculation of the power load factor of each target area based on the power demand forecast result with the highest matching degree in each target area specifically includes: Based on the power consumption data of each period in the time series data of the forecast result and the total number of the forecast period, the average load and maximum load in the period are obtained; The calculated average load and maximum load are input into the load factor formula to obtain the power load factor; Calculate the power load factor for each target area and obtain the power load factor time series dataset.

5. The method for optimizing power resource scheduling and load balancing based on big data according to claim 1, characterized in that: The construction of a regional load difference matrix based on the power load coefficient of each target area and the acquisition of a set of power resource transfer-out areas and a set of power resource transfer-in areas specifically include: Based on the obtained power load factor time series dataset, calculate the comprehensive load mean of all target areas; Based on the comprehensive load mean, calculate the difference between the power load of each target area and the comprehensive load mean, and construct a regional load difference matrix; Based on the difference values in the regional load difference matrix, a difference judgment threshold is set to divide and identify the outgoing and incoming areas; Based on the degree of difference, the transferred-out areas and the transferred-in areas are ranked, with the transferred-in areas represented by positive numbers and the transferred-out areas by negative numbers, and a scheduling docking priority sequence is established.

6. The method for optimizing power resource scheduling and load balancing based on big data according to claim 1, characterized in that: The method of dividing the scheduling period into multiple time periods based on the obtained outgoing region set and incoming region set, and performing power resource scheduling through incremental power transfer specifically includes: Divide the power dispatch cycle into hours, days, and weeks, and adjust it based on actual power demand; Based on the established dispatch docking priority sequence, sort the incoming regions from largest to smallest according to the difference value, and sort the outgoing regions from smallest to largest, and select the two target regions with the closest absolute difference value; Based on the divided dispatch period, the outgoing area will provide electricity during the period, and the incoming area will receive electricity during the period; According to the load difference and power demand changes, the power dispatch amount in each period is dynamically adjusted through the greedy algorithm; Real-time monitoring of power grid operation, actual load and dispatching effects in various regions is carried out, and the power demand forecasting model is retrained through the feedback mechanism to obtain stable forecasting results.

7. A method for optimizing power resource scheduling and load balancing based on big data, for implementing a system for optimizing power resource scheduling and load balancing based on big data as claimed in any one of claims 1 to 6, characterized in that: include: Data acquisition module: The data acquisition module is mainly responsible for collecting key operating data such as power consumption, voltage, current, frequency, etc. of power grid nodes in each target area in real time; Data processing module: The data processing module is mainly used to clean, normalize and extract features of historical electricity consumption data to provide standardized input for subsequent modeling; Power demand forecasting module: The power demand forecasting module is mainly used to make short-term and long-term forecasts of power demand in future time periods based on the LSTM model, and generate a set of forecast results under multiple confidence intervals; Similarity matching module: The similarity matching module is mainly used to perform dual matching of time series and spatial features between real-time collected data and prediction sets, and select the optimal prediction result under the current state; Load coefficient calculation module: The load coefficient calculation module is mainly used to calculate the average load and maximum load of each target area based on the selected prediction results, and obtain the load coefficient time series; Regional difference analysis module: The regional difference analysis module is mainly used to build a regional load difference matrix, identify the areas where power resources are transferred out and in, and establish a dispatch priority list based on load differences; Power resource scheduling module: The power resource scheduling module is mainly used to divide the scheduling cycle and use the greedy algorithm to formulate the optimal incremental power scheduling strategy to complete the transfer of power from the outgoing area to the incoming area; Feedback module: The feedback module is mainly used to evaluate the scheduling effect and update the power forecast model based on the scheduling results and real-time operation data; Processor: The processor is mainly used for the calculation process of each formula and the construction calculation process of each model.

8. An electronic device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the power resource scheduling and load balancing optimization method combined with big data as described in any one of claims 1-6.

9. A computer-readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, the power resource scheduling and load balancing optimization method combined with big data according to any one of claims 1 to 6 is implemented.

Images